On 1/19/2021 8:31 AM, Mark Shannon wrote:
It's time for yet another PEP :)
Fortunately, this one is a small one that doesn't change much.
It's aim is to make the VM more robust.
Abstract
========
This PEP proposes that machine stack overflow is treated differently
from runaway recursion.
My impression is that this is already the case in the sense given below.
https://bugs.python.org/issue42887 is about segfaults with this code
with no Python-level recursion.
mystr = "hello123"
for x in range(1000000):
mystr = mystr.__sizeof__()
Christian Heimes reproduced and said "The stack trace is several hundred
thousand (!) levels deep."
Ronald Oussoren said "The dealloc of "mystr" will cause recursive calls
to tp_dealloc along the entire chain and that can exhaust the C stack."
Since hundreds of thousands is a lot bigger than 1000, I assume that the
C recursion is not tracked. Whick is to say, recursive C calls are not
counted, unlike recursive Python calls. There have been several issues
about highly nested code and recursive structures causes segfaults,
presumably due to C level recursion. These are usually closed as 'Won't
fix' and the suggestion 'don't do that'.
I strongly agree that a fix would be great.
--
Terry Jan Reedy
_______________________________________________
Python-Dev mailing list -- python-dev@python.org
To unsubscribe send an email to python-dev-le...@python.org
https://mail.python.org/mailman3/lists/python-dev.python.org/
Message archived at
https://mail.python.org/archives/list/python-dev@python.org/message/4NA42RO7E5XV75WHWBEFVSJMFHJ7FCUJ/
Code of Conduct: http://python.org/psf/codeofconduct/
